G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS Analytica Chimica Acta xxx (2013) xxx–xxx
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density Uvaraj Uddayasankar, Ulrich J. Krull ∗ Chemical Sensors Group, Department of Chemical and Physical Sciences, University of Toronto Mississauga, 3359 Mississauga Road North, Mississauga, ON L5L 1C6, Canada
h i g h l i g h t s
g r a p h i c a l
a b s t r a c t
• Molecular beacons conjugated with surface immobilized gold nanoparticles. • Concurrent optical effects of nanoparticle immobilization density and size. • Nanoparticle size, density and spectral properties to optimize fluorescence signal.
a r t i c l e
i n f o
Article history: Received 28 April 2013 Received in revised form 28 June 2013 Accepted 27 July 2013 Available online xxx Keywords: Gold nanoparticles Molecular beacons Surface immobilization Optimization Fluorescence quenching
a b s t r a c t The high quenching efficiency of metal nanoparticles has facilitated its use as quenchers in molecular beacons. To optimize this system, a good understanding of the many factors that influence molecular beacon performance is required. In this study, molecular beacon performance was evaluated as a function of gold nanoparticle size and its immobilization characteristics. Gold nanoparticles of 4 nm, 15 nm and 87 nm diameter, were immobilized onto glass slides. Each size regime offered distinctive optical properties for fluorescence quenching of molecular dyes that were conjugated to oligonucleotides that were immobilized to the gold nanoparticles. Rigid double stranded DNA was used as a model to place fluorophores at different distances from the gold nanoparticles. The effect of particle size and also the immobilization density of nanoparticles was evaluated. The 4 nm and 87 nm gold nanoparticles offered the highest sensitivity in terms of the change in fluorescence intensity as a function of distance (3-fold improvement for Cy5). The optical properties of the molecular fluorophore was of significance, with Cy5 offering higher contrast ratios than Cy3 due to the red-shifted emission spectrum relative to the plasmon peak. A high density of gold nanoparticles reduced contrast ratios, indicating preference for a monolayer of immobilized nanoparticles when considering analytical performance. Molecular beacon probes were then used in place of the double stranded oligonucleotides. There was a strong dependence of molecular beacon performance on the length of a linker used for attachment to the nanoparticle surface. The optimal optical performance was obtained with 4 nm gold nanoparticles that were immobilized as monolayers of low density (5.7 × 1011 particles cm−2 ) on glass surfaces. These nanoparticle surfaces offered a 2-fold improvement in analytical performance of the molecular beacons when compared to other nanoparticle sizes investigated. The principles developed in this study would assist in the design of solid phase molecular beacons using gold nanoparticles. © 2013 Elsevier B.V. All rights reserved.
∗ Corresponding author. Tel.: +1 905 828 5437; fax: +1 905 828 5425. E-mail address: [email protected] (U.J. Krull). 0003-2670/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aca.2013.07.059
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
2
1. Introduction The highly efficient fluorescence quenching properties of gold nanoparticles has popularized their use in nucleic acid diagnostic schemes. A representative example of such a scheme is associated with the mechanism of function of molecular beacons [1]. The molecular beacons are constructed by immobilizing fluorophore labeled, hairpin shaped nucleic acid probes onto the nanoparticle surface. The hairpin shape holds the fluorophore in close proximity to the gold nanoparticle, effectively quenching fluorescence. Upon target hybridization, the fluorophore moves away from the nanoparticle, restoring fluorescence. Gold nanoparticles have been demonstrated to possess superior performance over molecular quenchers due to high extinction coefficients [1,2]. In addition, the broad plasmon peak of gold nanoparticles can quench multiple fluorophores, enabling multiplexed target detection [3]. Some effort has been directed to optimization of function of gold nanoparticle based molecular beacons in solution [4,5]. The investigation of gold nanoparticle based molecular beacons in solid phase assays has been limited. Moving this system to a solid support has a number of advantages [6]. Immobilizing the gold nanoparticles eliminates issues of colloidal stability. Bioconjugation, and subsequent washing steps, are facilitated as the nanoparticles are anchored on the solid substrate. The sensitivity of an assay can be improved by reducing the impact of the inner filter effect. The high extinction coefficient of nanoparticles makes them efficient quenchers, but it also serves to block the excitation radiation [7] and can reduce the sensitivity of bulk solution assays. We report herein the investigation of immobilized gold nanoparticles for optimization of the performance of solid phase assays using molecular beacons. More specifically, efforts were directed towards understanding the effect of gold nanoparticle size and immobilization density on molecular beacon performance. The feasibility of using surface immobilized nanoparticles as quenchers for nucleic acid assays has been previously demonstrated. Li et al. [8] designed a microfluidic device for nucleic acid detection using surface immobilized 10 nm gold nanoparticles. Their assay involved the use of Cy3 and FITC labeled nucleic acids, where the dyes were held in close proximity to the nanoparticle surface via adsorptive interactions. Hybridization removed the fluorophore from the surface, restoring fluorescence. Recently, Obliosca et al. [9] determined the distance dependent quenching properties of Cy3 on 10 nm gold nanoparticles that were immobilized on a glass slide. Double stranded DNA was used as a rigid spacer to control the distance between the dye and the nanoparticle. Peng et al. [10] reported a comprehensive study of immobilized silver nanostructures for use with molecular beacons. Silver nanostructures were grown on a solid substrate, and subsequently used as quenchers for molecular beacon probes. They demonstrated an improved performance for molecular beacons on the nanostructured surface as compared to a planar gold film. The effect of nanoparticle size on molecular beacon performance was also investigated along with application of these silver nanostructures in a microfluidic platform [11]. These studies indicate that immobilized nanoparticles offer a number of advantages for nucleic acid sensing. An investigation of the effect of gold nanoparticle size and immobilization density on solid phase molecular beacon performance has not been conducted, yet is fundamental in establishing assays of optimal performance. Three different sizes of nanoparticles (4, 15 and 87 nm) were chosen to represent distinct regimes of gold nanoparticle optical properties. These gold nanoparticles were immobilized onto glass substrates, followed by the immobilization of fluorescently labeled rigid double stranded nucleic acids. The DNA hybrids were labeled either on the 3 end or the 5 end, positioning the dye either distal or proximal to the surface.
Table 1 Oligonucleotide sequences. Linear probes SMN1 probe SMN1 3 TGT SMN1 5 TGT SMN1 probe pT Molecular beacons SMN1 MB SMN1 TGT
5 -ATT TTG TCT GAA ACC CTG T-3 – C6-Thiol F – 3 -TAA AAC AGA CTT TGG GAC A-5 3 -TAA AAC AGA CTT TGG GAC A-5 – F 5 -ATT TTG TCT GAA ACC CTG T TTTT TTTTTT-3 – C6-Thiol F – 5 -CCGGC ATT TTG TCT GAA ACC CTG T GCCGGTTTTTTTTTT-3 – C6-Thiol 3 -TAA AAC AGA CTT TGG GAC A-5
TGT = target; F = fluorophore (Cyanine 3 (Cy3) or Cyanine 5 (Cy5)); MB = molecular beacon.
Differences in the fluorescence intensities between these positions were used to evaluate the distance dependent quenching properties of the various nanoparticles. In addition, two different fluorophores, Cy3 and Cy5, were used to ascertain the effects of the emission wavelength on gold nanoparticle quenching efficiency. Molecular beacon probes were then used in place of the double stranded oligonucleotides. The ultimate goal of this study is to contribute towards a set of guidelines that may be used to design surface based molecular beacons using gold nanoparticles. These solid phase assays are increasing in importance due to the prevalence of microfluidic devices in analytical chemistry. The high surface area to volume ratio of these microsystems improve the sensitivity of these assays, and the characteristics of microfluidic flow also offer a number of advantages for solid phase assays [12]. 2. Experimental 2.1. Materials Ammonium hydroxide (30%), hydrochloric acid (12 M) and hydrogen peroxide (30%) were from EMD Chemicals (San Diego, CA). Methanol, tetrahydrofuran and isopropanol were at least reagent grade and obtained from Caledon Laboratories (Georgetown, ON, CA). Anhydrous dimethylformamide (DMF), (3Aminopropyl) trimethoxysilane (APTMS, 97%), lipoic acid (>98%), N,N -diisopropylcarbodiimide (DIC, 99%), N-hydroxysuccinimide (NHS, 98%), N,N-diisopropylethylamine (DIPEA, 99%), gold(III) chloride trihydrate (>99.9%), sodium citrate tribasic dehydrate (>99%), sodium borohydride (>96%) and hydroquinone (>99%) were from Sigma–Aldrich (Oakville, ON, CA). All buffers and aqueous solutions were prepared with deionized water from a Milli-Q cartridge purification system (Millipore Corporation, Mississauga, ON, CA). Water was sterilized by autoclaving. Autoclaved Tris borate buffer (TB, 50 mM, pH 7.4) with 1 M sodium chloride was used for all experiments. HPLC purified oligonucleotides (Table 1) were obtained from Integrated DNA Technologies (Coralville, IA). Glass microscope slides (75 mm × 25 mm) were from Fisher Scientific (Pittsburgh, PA). 2.2. Instrumentation UV–vis absorbance spectroscopy was done using an HP8452A diode array spectrophotometer (Hewlett Packard Corporation, Palo Alto, CA). Fluorescence spectra was obtained using a QuantaMaster Photon Technology International spectrofluorimeter (London, ON, Canada) with a xenon arc lamp (Ushio, Cypress, CA) as the excitation source and a red-sensitive R928P photomultiplier tube (Hamamatsu, Bridgewater, NJ). SEM images were obtained using an Hitachi S-5200 SEM (Hitachi High Technologies, America, Pleasanton, CA).
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
Epifluorescence microscopy was used to measure fluorescence intensities on the various surfaces. For Cy3, a Nikon C2si confocal microscope was used with the pinhole fully opened (i.e. epifluorescence mode). A 532 nm HeNe laser (1.5 mW) was used as the excitation source and a PMT was used as the detector. The excitation light was passed through a filter cube that consisted of a 530–560 nm bandpass excitation filter, a 570 nm longpass dichroic and a 573–648 nm bandpass emission filter. For Cy5, a Nikon Eclipse L150 epifluoroescence microscope was used. The excitation source was a 635 nm diode laser (10 mW). The intensity was controlled using neutral density filters, and the emitted fluorescence was monitored using a PMT. The excitation light was passed through a filter cube that consisted of a 630–650 nm bandpass excitation filter, a 660 nm longpass dichroic and a 665–695 nm bandpass emission filter. The images obtained were background corrected and quantified using ImageJ (NIH, USA) 2.3. Procedures 2.3.1. Glass slide preparation The RCA cleaning protocol [13] (Caution required) was used to treat the glass slides prior to silanization. The glass slides were immersed in a solution of ammonium hydroxide, hydrogen peroxide and water (1:1:5), which was heated to 80 ◦ C. After 10 min, they were rinsed copiously with water. The slides were then immersed in a solution of hydrochloric acid, hydrogen peroxide and water (1:1:5), which was heated to 80 ◦ C. After 10 min, the slides were rinsed copiously with water and dried under a stream of nitrogen. The slides were further dried in a vacuum oven at 100 ◦ C overnight. The slides were modified with a monolayer of APTMS following the protocol of Nath et al. [14]. Briefly, the dry glass slides were immersed in a methanolic solution of APTMS (5%, v/v) for 30 min with continuous agitation in an orbital shaker. The slides were then sonicated (water bath sonication) 4 times in methanol for 1 min per wash. The slides were then cured in a vacuum oven set at 120 ◦ C for 2 h. The resulting amine coated slides were immediately modified using the NHS ester of lipoic acid. The NHS ester of lipoic acid (LA-NHS) was synthesized using the protocol described by Algar et al. [15]. LA-NHS (0.1 g, 0.33 mmol) was dissolved in anhydrous DMF (60 mL) along with DIPEA (0.1 mL, 0.57 mmol). The amine functionalized glass slides were immersed in this solution and allowed to agitate in an orbital shaker for 24 h at room temperature. The slides were then rinsed at least 3 times with DMF using a water bath sonicator. The slides were subsequently rinsed twice with methanol, dried under a stream of nitrogen and stored in a vacuum desiccator. 2.3.2. Gold nanoparticle synthesis All glassware was cleaned with aqua regia prior to use (Caution required). Concentrated hydrochloric acid was mixed with nitric acid in a 3:1 (v/v) ratio. All glassware was rinsed with this solution for 10 min, followed by rinsing with deionized water. The glassware was dried overnight in an oven at 120 ◦ C. The 4 nm gold nanoparticles were synthesized as per the protocol of Jana et al. [16]. Briefly, gold(III) chloride (9.85 mg, 25 mol) was added to 100 mL of deionized water, followed by the addition of sodium citrate (7.35 mg, 25 mol). Sodium borohydride (9.5 mg, 250 mol) was then added with vigorous stirring. The solution was allowed to stir overnight, and then stored in a dark glass bottle at 4 ◦ C. The 15 nm gold nanoparticles were synthesized as per the protocol of Liu et al. [17]. Briefly, gold(III) chloride (39.4 mg, 0.1 mmol) was added to 100 mL of deionized water. This solution was brought to a vigorous boil, followed by the addition of sodium citrate (0.115 g, 0.39 mmol). The solution was allowed to boil for 20 min.
3
The solution was gently stirred overnight, and then stored in a dark glass bottle at 4 ◦ C. The 87 nm gold nanoparticles were prepared as per the protocol of Perrault et al. [18]. Briefly, 100 L of 15 nm gold nanoparticle solution (14.4 nM) was added to 50 mL of deionized water. Gold(III) chloride (5.79 mg, 14.7 mol) was added to this solution followed by the addition of sodium citrate (1.1 mg, 3.74 mol). Immediately after the addition of citrate, hydroquinone (1.65 mg, 15 mol) was rapidly added to the solution, which was vigorously stirred. The solution was gently stirred overnight, and then stored in a dark glass bottle at 4 ◦ C. 2.3.3. Gold nanoparticle immobilization The functionalized glass slides were rinsed with water, followed by immersion into a solution containing the gold nanoparticles. The concentration of the 4 nm gold nanoparticles was 0.1 M, the 15 nm gold nanoparticle solution was 11 nM and the 87 nm gold nanoparticle solution was 10 pM. The slides were allowed to react for at least 24 h with stirring by an orbital shaker to ensure reproducible immobilization. The slides were then rinsed with deionized water and stored in deionized water. 2.3.4. Rigid double stranded oligonucleotides To hybridize the DNA strands in solution, a 1 M solution of the 3 thiol functionalized probe was prepared in TB buffer, followed by the addition of 1.1 equivalents of the fluorescently labeled complementary strand. This solution was heated to 95 ◦ C for 5 min, and then allowed to reach room temperature over 1 h. Prior to nucleic acid functionalization, the gold nanoparticle coated slides were dried using a stream of nitrogen. Immobilization of the dsDNA was performed by incubating the nanoparticle surfaces with the hybridized DNA for 2 h, followed by rinsing with TB buffer. The substrates were scanned using the fluorescence microscopes. 2.3.5. Molecular beacon probes The molecular beacon probes were diluted to 1 M in TB buffer. For experiments that included mercaptohexanol as a diluent, this was added in the required ratio immediately prior to spotting solutions on the immobilized gold nanoparticles. The spotted solutions were allowed to incubate at room temperature for 2 h, followed by rinsing in boiling deionized water. For experiments where mercaptohexanol was backfilled onto the substrates, the substrates were incubated with a 1 mM mercaptohexanol solution in deionized water for 20 min before immersing the substrates in boiling water for rinsing. These substrates were then incubated in TB buffer for 1 h prior to scanning using a fluorescence microscope. Target hybridization was conducted with a concentration of 1 M target dissolved in TB buffer. To determine the concentration dependent response of the immobilized molecular beacons, the substrates were exposed to fully complementary targets at varying concentrations (dissolved in TB buffer). The substrates were incubated for 2 h, after which the fluorescence intensities were measured using an epifluorescence microscope. 2.3.6. Determination of molecular beacon loading capacity and hybridization efficiency DNA loading densities and hybridization efficiencies were determined using the procedure introduced by Demers et al. [19], with slight modifications. Probe immobilization procedures are identical to that described in Section 2.3.5. To displace the immobilized DNA, a 0.25 M dithiothreitol solution in TB buffer was used. The substrates were incubated with this solution for 12 h and the fluorescence intensity measured to determine the amount of DNA displaced. A calibration curve was constructed using solutions that
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
4
Table 2 Gold nanoparticle size and immobilization density derived from SEM images (Fig. 1). A total of 50 nanoparticles were used to determine the size and dispersity of nanoparticle size. Size of gold nanoparticle (nm)
Density of immobilized particles (particles cm−2 )
4.1 ± 0.7 15 ± 2 87 ± 12
5.7 × 1011 8 × 1010 7.3 × 108
contained known amounts of molecular beacon probe, with a solution composition identical to the displacement solution. To determine the amount of target hybridized, a fully complementary target strand, labeled with Cy5, was hybridized onto a surface that consisted of molecular beacons with a Cy3 label. An excess of target (10 M) and long hybridization times (12 h) were used to ensure saturation of hybridization. The substrates were rinsed with TB buffer followed by the displacement of the hybrids from the nanoparticle surface using a 0.25 M dithiothreitol solution in TB buffer. A calibration curve was constructed using the Cy5 labeled target in the same solution composition as the displacement solution. The amount of target hybridized was determined using fluorescence measurements. 3. Results and discussion 3.1. Surface immobilization of gold nanoparticles Three different sizes of gold nanoparticles, each belonging to a distinct optical regime, were chosen to investigate the effect of nanoparticle size on solid phase molecular beacon performance. One size regime involves nanoparticles that are less than 5 nm in diameter. Gold nanoparticles of this size scale have a damped plasmon resonance peak due to the reduced mean free path of the electrons [20]. The next size regime includes nanoparticles that are greater than 10 nm but less than 50 nm in diameter. Nanoparticles in this size range are known to have high fluorescence quenching efficiencies as their extinction spectrum is composed of an absorption component and no scattering coefficient [21]. The final size range represents nanoparticles that are greater than 60 nm in diameter. The extinction spectra of these nanoparticles have a significant scattering component in addition to an absorption component [21]. The sizes of the nanoparticles that were synthesized to represent each regime are given in Table 2. The absorbance spectra of the different gold nanoparticles can be found in the Supporting Information (Fig. S1). The gold nanoparticles were immobilized onto glass substrates using a bidentate, thiol based ligand (Supporting Information; Scheme 1). This involved the modification of glass substrates with (3-aminopropyl) trimethoxysilane to provide primary amine groups, which were then linked to lipoic acid by the formation of an amide bond. Using ellipsometry, the thickness of the APTMS layer was measured to be 0.8 nm, which was consistent with a monolayer of silane [22]. The surface modifications were confirmed using X-ray photoelectron spectroscopy (see Supporting Information, Table S1). While APTMS functionalized glass slides have been used for gold nanoparticle immobilization, the subsequent use in DNA assays would be problematic due to the possibility of DNA adsorption via electrostatic interactions. By blocking the amine functional group with lipoic acid, the adsorption of DNA can be minimized while still allowing gold nanoparticle immobilization. The gold nanoparticles immobilized onto the glass substrates were characterized by UV–vis spectroscopy and scanning electron microscopy (Fig. 1). The absorption spectra of the immobilized gold nanoparticles were red-shifted as compared to that in solution, but the width of the plasmon peak was not significantly increased. This
is indicative of a monolayer of nanoparticles that are spaced far apart to minimize plasmonic coupling [14]. The red shift in the plasmon peak can be attributed to the higher refractive index of the glass substrate and minimal plasmonic coupling due to the higher packing density of nanoparticles on a surface, as compared to solution. SEM imaging was used to confirm gold nanoparticle size and the surface density of the nanoparticles. Table 2 summarizes the information derived from the SEM images. The particle immobilization density decreased as the size of the nanoparticle increased. There are two contributing factors for this observation, namely the charge on the nanoparticles and the nanoparticle concentration. Immobilization density is controlled by the electrostatic repulsion between the nanoparticles, and the affinity of the nanoparticle for the anchoring ligand on the substrate. All sizes of nanoparticles had citrate as a capping ligand, making them negatively charged. The greater surface area of the large nanoparticles results in a higher negative charge. This leads to stronger repulsion between the larger nanoparticles resulting in a lower immobilization density. These results are in accordance with those of Nath et al. [14], in which nanoparticles, ranging from 12 nm to 50 nm, were immobilized onto APTMS surfaces. They observed a lower immobilization density for the larger nanoparticles. Another factor influencing surface coverage of the nanoparticles is the concentration of nanoparticles in the deposition solution. The stock solutions of the larger nanoparticles were much lower compared to the smaller nanoparticles. Lower concentrations are known to result in a lower surface coverage, as was previously observed by Keating et al. [23]. Attempts to increase the concentration of the citrate coated gold nanoparticles resulted in irreversible aggregation, so the solutions were used as synthesized. Incubation times greater than 24 h were used to ensure saturation of nanoparticle immobilization, even for the low concentrations of gold nanoparticles. 3.2. Distance dependent quenching properties of immobilized gold nanoparticles Variations in nanoparticle size can be expected to influence molecular beacon performance via two mechanisms. First, the packing density of nucleic acids on the surface of a nanoparticle is known to vary significantly as a function of nanoparticle size [24]. This in turn could affect the conformation of the molecular beacon, resulting in changes in the fluorescence signal observed from the fluorophores. The other aspect is the variations in the distance dependent quenching efficiency as a function of nanoparticle size. To isolate the optical effects from the physical aspects, fluorescently labeled double stranded DNA was used for initial experiments. More specifically, a 3 -thiol modified probe (19-mer) was hybridized with targets that were labeled with a fluorophore on either the 3 or 5 end (Fig. 2). Upon immobilization on the nanoparticles, the 5 position places the dye proximal to the gold surface (approx. 1.1 nm), while the 3 end locates the dye distal (approx. 8 nm) to the gold surface. The proximal label emulates a scenario where all the molecular beacons are in the closed conformation (i.e. no hybridization). The distal label emulates the situation where all the molecular beacons are hybridized to the target (i.e. 100% hybridization). The ratio of fluorescence intensities between the two positions (the contrast ratio) was used as a surrogate to gauge molecular beacon performance. Pre-hybridizing provided for linear probes prior to immobilization, and eliminated the effect of varying hybridization efficiency as a function of nanoparticle size. In addition, double stranded DNA is less prone to non-specific adsorption onto gold surfaces, eliminating the effect of surface interactions on probe conformation. The optical properties of the two different hybrids were confirmed to
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
5
Fig. 1. UV–vis spectra and SEM analysis of surface immobilized gold nanoparticles: (a) 4 nm gold nanoparticles, (b) 15 nm gold nanoparticles and (c) 87 nm gold nanoparticles. The UV–vis spectra compare the optical properties of the gold nanoparticles in solution (solid) and immobilized on a glass substrate (dotted) in an aqueous environment. The SEM images were obtained on silicon wafers that were chemically modified with lipoic acid.
be identical (see Supporting Information, Fig. S2). This verification was important because it is known that quantum yields of certain fluorophores are affected by the position in a nucleic acid sequence [25]. Differences in fluorescence intensity between the distal and proximal positions were measured for two dyes, Cy3 and Cy5, on three different sizes of gold nanoparticles immobilized on solid substrates. The fluorescence intensities from these experiments are reported in Fig. 3. Molecular beacon performance is characterized by the change in fluorescence intensity upon hybridization. It is therefore appropriate to consider contrast ratios rather than absolute intensities. For clarity, contrast ratio refers to the ratio of fluorescence intensity of the distal positioned dye to the proximal positioned dye. Comparing the Cy3 contrast ratios (Fig. 3a) on the 4 nm and 15 nm gold nanoparticles, the smaller nanoparticles have a 1.6-fold greater contrast ratio, indicating potentially better molecular beacon performance. The distance dependent quenching for nanoparticles smaller than 60 nm is known to be accurately modeled using
Nanomaterial Surface Energy Transfer (NSET) theory [26]. From NSET theory, the 4 nm gold nanoparticles have a quenching efficiency of more than 95% at distances less than 2 nm from the surface. The distance between the proximal label and the nanoparticle surface falls below this limit, indicating a good quenching efficiency even for the small gold nanoparticles. As the nanoparticle size increases, the quenching efficiency is expected to increase due to higher extinction coefficients. Thus, the quenching efficiency for the 15 nm gold nanoparticles is expected to be higher than the 4 nm gold nanoparticles. The distal positioned fluorophore was expected to be approximately 7.8 nm from the nanoparticle surface. At these distances, Chhabra et al. [26] observed that 5 nm gold nanoparticles had a quenching efficiency of 60% for a Cy3 dye. In the same study, 10 nm gold nanoparticles had a 90% quenching efficiency for the same position of fluorophore. For the small gold nanoparticles, the damped plasmon peak results in a lower quenching efficiency when compared to larger gold nanoparticles. This allows the smaller gold nanoparticles to have a higher fluorescence intensity at the distal position, but effectively quench fluorescence when the dye is
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10 6
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
Fig. 2. Illustration of the fluorophore position for the 3 and 5 labeled hybrids that are immobilized on gold nanoparticle surfaces to determine the distance dependent quenching efficiency.
proximal to the surface, leading to higher contrast ratios for the 4 nm gold nanoparticles Comparing the contrast ratios for the 15 nm and the 87 nm gold nanoparticles, the larger gold nanoparticles offer a 1.3-fold greater performance. Larger nanoparticles would be expected to quench fluorescence more efficiently, but the greater scattering coefficient of the large nanoparticles could also enhance the fluorescence. The balance of these two factors would determine the contrast ratio. In a recent study of molecular beacon performance reported by
Cheng et al. [5], the use of 100 nm gold nanoparticles was compared to 20 and 50 nm gold nanoparticles. They demonstrated that the quenching efficiency at close proximity to the surface was similar for gold nanoparticle sizes greater than 20 nm. When spaced 10 nm away from the surface, the 20 nm gold nanoparticles offered less than 20% of an increase in fluorescence intensity, while the larger gold nanoparticles (100 nm) offered greater than 60% increase in fluorescence. Our results are in agreement with their observations, with the larger gold nanoparticles offering a greater contrast ratio. When Cy5 was used as the fluorophore, the trend in the contrast ratio as a function of nanoparticle size (Fig. 3b) was similar to that of Cy3. While the trend was similar, the contrast ratios for the 4 and 87 nm gold nanoparticles were 3-fold greater than the 15 nm gold nanoparticles (compared to
ARTICLE IN PRESS Analytica Chimica Acta xxx (2013) xxx–xxx
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density Uvaraj Uddayasankar, Ulrich J. Krull ∗ Chemical Sensors Group, Department of Chemical and Physical Sciences, University of Toronto Mississauga, 3359 Mississauga Road North, Mississauga, ON L5L 1C6, Canada
h i g h l i g h t s
g r a p h i c a l
a b s t r a c t
• Molecular beacons conjugated with surface immobilized gold nanoparticles. • Concurrent optical effects of nanoparticle immobilization density and size. • Nanoparticle size, density and spectral properties to optimize fluorescence signal.
a r t i c l e
i n f o
Article history: Received 28 April 2013 Received in revised form 28 June 2013 Accepted 27 July 2013 Available online xxx Keywords: Gold nanoparticles Molecular beacons Surface immobilization Optimization Fluorescence quenching
a b s t r a c t The high quenching efficiency of metal nanoparticles has facilitated its use as quenchers in molecular beacons. To optimize this system, a good understanding of the many factors that influence molecular beacon performance is required. In this study, molecular beacon performance was evaluated as a function of gold nanoparticle size and its immobilization characteristics. Gold nanoparticles of 4 nm, 15 nm and 87 nm diameter, were immobilized onto glass slides. Each size regime offered distinctive optical properties for fluorescence quenching of molecular dyes that were conjugated to oligonucleotides that were immobilized to the gold nanoparticles. Rigid double stranded DNA was used as a model to place fluorophores at different distances from the gold nanoparticles. The effect of particle size and also the immobilization density of nanoparticles was evaluated. The 4 nm and 87 nm gold nanoparticles offered the highest sensitivity in terms of the change in fluorescence intensity as a function of distance (3-fold improvement for Cy5). The optical properties of the molecular fluorophore was of significance, with Cy5 offering higher contrast ratios than Cy3 due to the red-shifted emission spectrum relative to the plasmon peak. A high density of gold nanoparticles reduced contrast ratios, indicating preference for a monolayer of immobilized nanoparticles when considering analytical performance. Molecular beacon probes were then used in place of the double stranded oligonucleotides. There was a strong dependence of molecular beacon performance on the length of a linker used for attachment to the nanoparticle surface. The optimal optical performance was obtained with 4 nm gold nanoparticles that were immobilized as monolayers of low density (5.7 × 1011 particles cm−2 ) on glass surfaces. These nanoparticle surfaces offered a 2-fold improvement in analytical performance of the molecular beacons when compared to other nanoparticle sizes investigated. The principles developed in this study would assist in the design of solid phase molecular beacons using gold nanoparticles. © 2013 Elsevier B.V. All rights reserved.
∗ Corresponding author. Tel.: +1 905 828 5437; fax: +1 905 828 5425. E-mail address: [email protected] (U.J. Krull). 0003-2670/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aca.2013.07.059
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
2
1. Introduction The highly efficient fluorescence quenching properties of gold nanoparticles has popularized their use in nucleic acid diagnostic schemes. A representative example of such a scheme is associated with the mechanism of function of molecular beacons [1]. The molecular beacons are constructed by immobilizing fluorophore labeled, hairpin shaped nucleic acid probes onto the nanoparticle surface. The hairpin shape holds the fluorophore in close proximity to the gold nanoparticle, effectively quenching fluorescence. Upon target hybridization, the fluorophore moves away from the nanoparticle, restoring fluorescence. Gold nanoparticles have been demonstrated to possess superior performance over molecular quenchers due to high extinction coefficients [1,2]. In addition, the broad plasmon peak of gold nanoparticles can quench multiple fluorophores, enabling multiplexed target detection [3]. Some effort has been directed to optimization of function of gold nanoparticle based molecular beacons in solution [4,5]. The investigation of gold nanoparticle based molecular beacons in solid phase assays has been limited. Moving this system to a solid support has a number of advantages [6]. Immobilizing the gold nanoparticles eliminates issues of colloidal stability. Bioconjugation, and subsequent washing steps, are facilitated as the nanoparticles are anchored on the solid substrate. The sensitivity of an assay can be improved by reducing the impact of the inner filter effect. The high extinction coefficient of nanoparticles makes them efficient quenchers, but it also serves to block the excitation radiation [7] and can reduce the sensitivity of bulk solution assays. We report herein the investigation of immobilized gold nanoparticles for optimization of the performance of solid phase assays using molecular beacons. More specifically, efforts were directed towards understanding the effect of gold nanoparticle size and immobilization density on molecular beacon performance. The feasibility of using surface immobilized nanoparticles as quenchers for nucleic acid assays has been previously demonstrated. Li et al. [8] designed a microfluidic device for nucleic acid detection using surface immobilized 10 nm gold nanoparticles. Their assay involved the use of Cy3 and FITC labeled nucleic acids, where the dyes were held in close proximity to the nanoparticle surface via adsorptive interactions. Hybridization removed the fluorophore from the surface, restoring fluorescence. Recently, Obliosca et al. [9] determined the distance dependent quenching properties of Cy3 on 10 nm gold nanoparticles that were immobilized on a glass slide. Double stranded DNA was used as a rigid spacer to control the distance between the dye and the nanoparticle. Peng et al. [10] reported a comprehensive study of immobilized silver nanostructures for use with molecular beacons. Silver nanostructures were grown on a solid substrate, and subsequently used as quenchers for molecular beacon probes. They demonstrated an improved performance for molecular beacons on the nanostructured surface as compared to a planar gold film. The effect of nanoparticle size on molecular beacon performance was also investigated along with application of these silver nanostructures in a microfluidic platform [11]. These studies indicate that immobilized nanoparticles offer a number of advantages for nucleic acid sensing. An investigation of the effect of gold nanoparticle size and immobilization density on solid phase molecular beacon performance has not been conducted, yet is fundamental in establishing assays of optimal performance. Three different sizes of nanoparticles (4, 15 and 87 nm) were chosen to represent distinct regimes of gold nanoparticle optical properties. These gold nanoparticles were immobilized onto glass substrates, followed by the immobilization of fluorescently labeled rigid double stranded nucleic acids. The DNA hybrids were labeled either on the 3 end or the 5 end, positioning the dye either distal or proximal to the surface.
Table 1 Oligonucleotide sequences. Linear probes SMN1 probe SMN1 3 TGT SMN1 5 TGT SMN1 probe pT Molecular beacons SMN1 MB SMN1 TGT
5 -ATT TTG TCT GAA ACC CTG T-3 – C6-Thiol F – 3 -TAA AAC AGA CTT TGG GAC A-5 3 -TAA AAC AGA CTT TGG GAC A-5 – F 5 -ATT TTG TCT GAA ACC CTG T TTTT TTTTTT-3 – C6-Thiol F – 5 -CCGGC ATT TTG TCT GAA ACC CTG T GCCGGTTTTTTTTTT-3 – C6-Thiol 3 -TAA AAC AGA CTT TGG GAC A-5
TGT = target; F = fluorophore (Cyanine 3 (Cy3) or Cyanine 5 (Cy5)); MB = molecular beacon.
Differences in the fluorescence intensities between these positions were used to evaluate the distance dependent quenching properties of the various nanoparticles. In addition, two different fluorophores, Cy3 and Cy5, were used to ascertain the effects of the emission wavelength on gold nanoparticle quenching efficiency. Molecular beacon probes were then used in place of the double stranded oligonucleotides. The ultimate goal of this study is to contribute towards a set of guidelines that may be used to design surface based molecular beacons using gold nanoparticles. These solid phase assays are increasing in importance due to the prevalence of microfluidic devices in analytical chemistry. The high surface area to volume ratio of these microsystems improve the sensitivity of these assays, and the characteristics of microfluidic flow also offer a number of advantages for solid phase assays [12]. 2. Experimental 2.1. Materials Ammonium hydroxide (30%), hydrochloric acid (12 M) and hydrogen peroxide (30%) were from EMD Chemicals (San Diego, CA). Methanol, tetrahydrofuran and isopropanol were at least reagent grade and obtained from Caledon Laboratories (Georgetown, ON, CA). Anhydrous dimethylformamide (DMF), (3Aminopropyl) trimethoxysilane (APTMS, 97%), lipoic acid (>98%), N,N -diisopropylcarbodiimide (DIC, 99%), N-hydroxysuccinimide (NHS, 98%), N,N-diisopropylethylamine (DIPEA, 99%), gold(III) chloride trihydrate (>99.9%), sodium citrate tribasic dehydrate (>99%), sodium borohydride (>96%) and hydroquinone (>99%) were from Sigma–Aldrich (Oakville, ON, CA). All buffers and aqueous solutions were prepared with deionized water from a Milli-Q cartridge purification system (Millipore Corporation, Mississauga, ON, CA). Water was sterilized by autoclaving. Autoclaved Tris borate buffer (TB, 50 mM, pH 7.4) with 1 M sodium chloride was used for all experiments. HPLC purified oligonucleotides (Table 1) were obtained from Integrated DNA Technologies (Coralville, IA). Glass microscope slides (75 mm × 25 mm) were from Fisher Scientific (Pittsburgh, PA). 2.2. Instrumentation UV–vis absorbance spectroscopy was done using an HP8452A diode array spectrophotometer (Hewlett Packard Corporation, Palo Alto, CA). Fluorescence spectra was obtained using a QuantaMaster Photon Technology International spectrofluorimeter (London, ON, Canada) with a xenon arc lamp (Ushio, Cypress, CA) as the excitation source and a red-sensitive R928P photomultiplier tube (Hamamatsu, Bridgewater, NJ). SEM images were obtained using an Hitachi S-5200 SEM (Hitachi High Technologies, America, Pleasanton, CA).
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
Epifluorescence microscopy was used to measure fluorescence intensities on the various surfaces. For Cy3, a Nikon C2si confocal microscope was used with the pinhole fully opened (i.e. epifluorescence mode). A 532 nm HeNe laser (1.5 mW) was used as the excitation source and a PMT was used as the detector. The excitation light was passed through a filter cube that consisted of a 530–560 nm bandpass excitation filter, a 570 nm longpass dichroic and a 573–648 nm bandpass emission filter. For Cy5, a Nikon Eclipse L150 epifluoroescence microscope was used. The excitation source was a 635 nm diode laser (10 mW). The intensity was controlled using neutral density filters, and the emitted fluorescence was monitored using a PMT. The excitation light was passed through a filter cube that consisted of a 630–650 nm bandpass excitation filter, a 660 nm longpass dichroic and a 665–695 nm bandpass emission filter. The images obtained were background corrected and quantified using ImageJ (NIH, USA) 2.3. Procedures 2.3.1. Glass slide preparation The RCA cleaning protocol [13] (Caution required) was used to treat the glass slides prior to silanization. The glass slides were immersed in a solution of ammonium hydroxide, hydrogen peroxide and water (1:1:5), which was heated to 80 ◦ C. After 10 min, they were rinsed copiously with water. The slides were then immersed in a solution of hydrochloric acid, hydrogen peroxide and water (1:1:5), which was heated to 80 ◦ C. After 10 min, the slides were rinsed copiously with water and dried under a stream of nitrogen. The slides were further dried in a vacuum oven at 100 ◦ C overnight. The slides were modified with a monolayer of APTMS following the protocol of Nath et al. [14]. Briefly, the dry glass slides were immersed in a methanolic solution of APTMS (5%, v/v) for 30 min with continuous agitation in an orbital shaker. The slides were then sonicated (water bath sonication) 4 times in methanol for 1 min per wash. The slides were then cured in a vacuum oven set at 120 ◦ C for 2 h. The resulting amine coated slides were immediately modified using the NHS ester of lipoic acid. The NHS ester of lipoic acid (LA-NHS) was synthesized using the protocol described by Algar et al. [15]. LA-NHS (0.1 g, 0.33 mmol) was dissolved in anhydrous DMF (60 mL) along with DIPEA (0.1 mL, 0.57 mmol). The amine functionalized glass slides were immersed in this solution and allowed to agitate in an orbital shaker for 24 h at room temperature. The slides were then rinsed at least 3 times with DMF using a water bath sonicator. The slides were subsequently rinsed twice with methanol, dried under a stream of nitrogen and stored in a vacuum desiccator. 2.3.2. Gold nanoparticle synthesis All glassware was cleaned with aqua regia prior to use (Caution required). Concentrated hydrochloric acid was mixed with nitric acid in a 3:1 (v/v) ratio. All glassware was rinsed with this solution for 10 min, followed by rinsing with deionized water. The glassware was dried overnight in an oven at 120 ◦ C. The 4 nm gold nanoparticles were synthesized as per the protocol of Jana et al. [16]. Briefly, gold(III) chloride (9.85 mg, 25 mol) was added to 100 mL of deionized water, followed by the addition of sodium citrate (7.35 mg, 25 mol). Sodium borohydride (9.5 mg, 250 mol) was then added with vigorous stirring. The solution was allowed to stir overnight, and then stored in a dark glass bottle at 4 ◦ C. The 15 nm gold nanoparticles were synthesized as per the protocol of Liu et al. [17]. Briefly, gold(III) chloride (39.4 mg, 0.1 mmol) was added to 100 mL of deionized water. This solution was brought to a vigorous boil, followed by the addition of sodium citrate (0.115 g, 0.39 mmol). The solution was allowed to boil for 20 min.
3
The solution was gently stirred overnight, and then stored in a dark glass bottle at 4 ◦ C. The 87 nm gold nanoparticles were prepared as per the protocol of Perrault et al. [18]. Briefly, 100 L of 15 nm gold nanoparticle solution (14.4 nM) was added to 50 mL of deionized water. Gold(III) chloride (5.79 mg, 14.7 mol) was added to this solution followed by the addition of sodium citrate (1.1 mg, 3.74 mol). Immediately after the addition of citrate, hydroquinone (1.65 mg, 15 mol) was rapidly added to the solution, which was vigorously stirred. The solution was gently stirred overnight, and then stored in a dark glass bottle at 4 ◦ C. 2.3.3. Gold nanoparticle immobilization The functionalized glass slides were rinsed with water, followed by immersion into a solution containing the gold nanoparticles. The concentration of the 4 nm gold nanoparticles was 0.1 M, the 15 nm gold nanoparticle solution was 11 nM and the 87 nm gold nanoparticle solution was 10 pM. The slides were allowed to react for at least 24 h with stirring by an orbital shaker to ensure reproducible immobilization. The slides were then rinsed with deionized water and stored in deionized water. 2.3.4. Rigid double stranded oligonucleotides To hybridize the DNA strands in solution, a 1 M solution of the 3 thiol functionalized probe was prepared in TB buffer, followed by the addition of 1.1 equivalents of the fluorescently labeled complementary strand. This solution was heated to 95 ◦ C for 5 min, and then allowed to reach room temperature over 1 h. Prior to nucleic acid functionalization, the gold nanoparticle coated slides were dried using a stream of nitrogen. Immobilization of the dsDNA was performed by incubating the nanoparticle surfaces with the hybridized DNA for 2 h, followed by rinsing with TB buffer. The substrates were scanned using the fluorescence microscopes. 2.3.5. Molecular beacon probes The molecular beacon probes were diluted to 1 M in TB buffer. For experiments that included mercaptohexanol as a diluent, this was added in the required ratio immediately prior to spotting solutions on the immobilized gold nanoparticles. The spotted solutions were allowed to incubate at room temperature for 2 h, followed by rinsing in boiling deionized water. For experiments where mercaptohexanol was backfilled onto the substrates, the substrates were incubated with a 1 mM mercaptohexanol solution in deionized water for 20 min before immersing the substrates in boiling water for rinsing. These substrates were then incubated in TB buffer for 1 h prior to scanning using a fluorescence microscope. Target hybridization was conducted with a concentration of 1 M target dissolved in TB buffer. To determine the concentration dependent response of the immobilized molecular beacons, the substrates were exposed to fully complementary targets at varying concentrations (dissolved in TB buffer). The substrates were incubated for 2 h, after which the fluorescence intensities were measured using an epifluorescence microscope. 2.3.6. Determination of molecular beacon loading capacity and hybridization efficiency DNA loading densities and hybridization efficiencies were determined using the procedure introduced by Demers et al. [19], with slight modifications. Probe immobilization procedures are identical to that described in Section 2.3.5. To displace the immobilized DNA, a 0.25 M dithiothreitol solution in TB buffer was used. The substrates were incubated with this solution for 12 h and the fluorescence intensity measured to determine the amount of DNA displaced. A calibration curve was constructed using solutions that
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
4
Table 2 Gold nanoparticle size and immobilization density derived from SEM images (Fig. 1). A total of 50 nanoparticles were used to determine the size and dispersity of nanoparticle size. Size of gold nanoparticle (nm)
Density of immobilized particles (particles cm−2 )
4.1 ± 0.7 15 ± 2 87 ± 12
5.7 × 1011 8 × 1010 7.3 × 108
contained known amounts of molecular beacon probe, with a solution composition identical to the displacement solution. To determine the amount of target hybridized, a fully complementary target strand, labeled with Cy5, was hybridized onto a surface that consisted of molecular beacons with a Cy3 label. An excess of target (10 M) and long hybridization times (12 h) were used to ensure saturation of hybridization. The substrates were rinsed with TB buffer followed by the displacement of the hybrids from the nanoparticle surface using a 0.25 M dithiothreitol solution in TB buffer. A calibration curve was constructed using the Cy5 labeled target in the same solution composition as the displacement solution. The amount of target hybridized was determined using fluorescence measurements. 3. Results and discussion 3.1. Surface immobilization of gold nanoparticles Three different sizes of gold nanoparticles, each belonging to a distinct optical regime, were chosen to investigate the effect of nanoparticle size on solid phase molecular beacon performance. One size regime involves nanoparticles that are less than 5 nm in diameter. Gold nanoparticles of this size scale have a damped plasmon resonance peak due to the reduced mean free path of the electrons [20]. The next size regime includes nanoparticles that are greater than 10 nm but less than 50 nm in diameter. Nanoparticles in this size range are known to have high fluorescence quenching efficiencies as their extinction spectrum is composed of an absorption component and no scattering coefficient [21]. The final size range represents nanoparticles that are greater than 60 nm in diameter. The extinction spectra of these nanoparticles have a significant scattering component in addition to an absorption component [21]. The sizes of the nanoparticles that were synthesized to represent each regime are given in Table 2. The absorbance spectra of the different gold nanoparticles can be found in the Supporting Information (Fig. S1). The gold nanoparticles were immobilized onto glass substrates using a bidentate, thiol based ligand (Supporting Information; Scheme 1). This involved the modification of glass substrates with (3-aminopropyl) trimethoxysilane to provide primary amine groups, which were then linked to lipoic acid by the formation of an amide bond. Using ellipsometry, the thickness of the APTMS layer was measured to be 0.8 nm, which was consistent with a monolayer of silane [22]. The surface modifications were confirmed using X-ray photoelectron spectroscopy (see Supporting Information, Table S1). While APTMS functionalized glass slides have been used for gold nanoparticle immobilization, the subsequent use in DNA assays would be problematic due to the possibility of DNA adsorption via electrostatic interactions. By blocking the amine functional group with lipoic acid, the adsorption of DNA can be minimized while still allowing gold nanoparticle immobilization. The gold nanoparticles immobilized onto the glass substrates were characterized by UV–vis spectroscopy and scanning electron microscopy (Fig. 1). The absorption spectra of the immobilized gold nanoparticles were red-shifted as compared to that in solution, but the width of the plasmon peak was not significantly increased. This
is indicative of a monolayer of nanoparticles that are spaced far apart to minimize plasmonic coupling [14]. The red shift in the plasmon peak can be attributed to the higher refractive index of the glass substrate and minimal plasmonic coupling due to the higher packing density of nanoparticles on a surface, as compared to solution. SEM imaging was used to confirm gold nanoparticle size and the surface density of the nanoparticles. Table 2 summarizes the information derived from the SEM images. The particle immobilization density decreased as the size of the nanoparticle increased. There are two contributing factors for this observation, namely the charge on the nanoparticles and the nanoparticle concentration. Immobilization density is controlled by the electrostatic repulsion between the nanoparticles, and the affinity of the nanoparticle for the anchoring ligand on the substrate. All sizes of nanoparticles had citrate as a capping ligand, making them negatively charged. The greater surface area of the large nanoparticles results in a higher negative charge. This leads to stronger repulsion between the larger nanoparticles resulting in a lower immobilization density. These results are in accordance with those of Nath et al. [14], in which nanoparticles, ranging from 12 nm to 50 nm, were immobilized onto APTMS surfaces. They observed a lower immobilization density for the larger nanoparticles. Another factor influencing surface coverage of the nanoparticles is the concentration of nanoparticles in the deposition solution. The stock solutions of the larger nanoparticles were much lower compared to the smaller nanoparticles. Lower concentrations are known to result in a lower surface coverage, as was previously observed by Keating et al. [23]. Attempts to increase the concentration of the citrate coated gold nanoparticles resulted in irreversible aggregation, so the solutions were used as synthesized. Incubation times greater than 24 h were used to ensure saturation of nanoparticle immobilization, even for the low concentrations of gold nanoparticles. 3.2. Distance dependent quenching properties of immobilized gold nanoparticles Variations in nanoparticle size can be expected to influence molecular beacon performance via two mechanisms. First, the packing density of nucleic acids on the surface of a nanoparticle is known to vary significantly as a function of nanoparticle size [24]. This in turn could affect the conformation of the molecular beacon, resulting in changes in the fluorescence signal observed from the fluorophores. The other aspect is the variations in the distance dependent quenching efficiency as a function of nanoparticle size. To isolate the optical effects from the physical aspects, fluorescently labeled double stranded DNA was used for initial experiments. More specifically, a 3 -thiol modified probe (19-mer) was hybridized with targets that were labeled with a fluorophore on either the 3 or 5 end (Fig. 2). Upon immobilization on the nanoparticles, the 5 position places the dye proximal to the gold surface (approx. 1.1 nm), while the 3 end locates the dye distal (approx. 8 nm) to the gold surface. The proximal label emulates a scenario where all the molecular beacons are in the closed conformation (i.e. no hybridization). The distal label emulates the situation where all the molecular beacons are hybridized to the target (i.e. 100% hybridization). The ratio of fluorescence intensities between the two positions (the contrast ratio) was used as a surrogate to gauge molecular beacon performance. Pre-hybridizing provided for linear probes prior to immobilization, and eliminated the effect of varying hybridization efficiency as a function of nanoparticle size. In addition, double stranded DNA is less prone to non-specific adsorption onto gold surfaces, eliminating the effect of surface interactions on probe conformation. The optical properties of the two different hybrids were confirmed to
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
5
Fig. 1. UV–vis spectra and SEM analysis of surface immobilized gold nanoparticles: (a) 4 nm gold nanoparticles, (b) 15 nm gold nanoparticles and (c) 87 nm gold nanoparticles. The UV–vis spectra compare the optical properties of the gold nanoparticles in solution (solid) and immobilized on a glass substrate (dotted) in an aqueous environment. The SEM images were obtained on silicon wafers that were chemically modified with lipoic acid.
be identical (see Supporting Information, Fig. S2). This verification was important because it is known that quantum yields of certain fluorophores are affected by the position in a nucleic acid sequence [25]. Differences in fluorescence intensity between the distal and proximal positions were measured for two dyes, Cy3 and Cy5, on three different sizes of gold nanoparticles immobilized on solid substrates. The fluorescence intensities from these experiments are reported in Fig. 3. Molecular beacon performance is characterized by the change in fluorescence intensity upon hybridization. It is therefore appropriate to consider contrast ratios rather than absolute intensities. For clarity, contrast ratio refers to the ratio of fluorescence intensity of the distal positioned dye to the proximal positioned dye. Comparing the Cy3 contrast ratios (Fig. 3a) on the 4 nm and 15 nm gold nanoparticles, the smaller nanoparticles have a 1.6-fold greater contrast ratio, indicating potentially better molecular beacon performance. The distance dependent quenching for nanoparticles smaller than 60 nm is known to be accurately modeled using
Nanomaterial Surface Energy Transfer (NSET) theory [26]. From NSET theory, the 4 nm gold nanoparticles have a quenching efficiency of more than 95% at distances less than 2 nm from the surface. The distance between the proximal label and the nanoparticle surface falls below this limit, indicating a good quenching efficiency even for the small gold nanoparticles. As the nanoparticle size increases, the quenching efficiency is expected to increase due to higher extinction coefficients. Thus, the quenching efficiency for the 15 nm gold nanoparticles is expected to be higher than the 4 nm gold nanoparticles. The distal positioned fluorophore was expected to be approximately 7.8 nm from the nanoparticle surface. At these distances, Chhabra et al. [26] observed that 5 nm gold nanoparticles had a quenching efficiency of 60% for a Cy3 dye. In the same study, 10 nm gold nanoparticles had a 90% quenching efficiency for the same position of fluorophore. For the small gold nanoparticles, the damped plasmon peak results in a lower quenching efficiency when compared to larger gold nanoparticles. This allows the smaller gold nanoparticles to have a higher fluorescence intensity at the distal position, but effectively quench fluorescence when the dye is
Please cite this article in press as: U. Uddayasankar, U.J. Krull, Analytical performance of molecular beacons on surface immobilized gold nanoparticles of varying size and density, Anal. Chim. Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.07.059
G Model ACA-232745; No. of Pages 10 6
ARTICLE IN PRESS U. Uddayasankar, U.J. Krull / Analytica Chimica Acta xxx (2013) xxx–xxx
Fig. 2. Illustration of the fluorophore position for the 3 and 5 labeled hybrids that are immobilized on gold nanoparticle surfaces to determine the distance dependent quenching efficiency.
proximal to the surface, leading to higher contrast ratios for the 4 nm gold nanoparticles Comparing the contrast ratios for the 15 nm and the 87 nm gold nanoparticles, the larger gold nanoparticles offer a 1.3-fold greater performance. Larger nanoparticles would be expected to quench fluorescence more efficiently, but the greater scattering coefficient of the large nanoparticles could also enhance the fluorescence. The balance of these two factors would determine the contrast ratio. In a recent study of molecular beacon performance reported by
Cheng et al. [5], the use of 100 nm gold nanoparticles was compared to 20 and 50 nm gold nanoparticles. They demonstrated that the quenching efficiency at close proximity to the surface was similar for gold nanoparticle sizes greater than 20 nm. When spaced 10 nm away from the surface, the 20 nm gold nanoparticles offered less than 20% of an increase in fluorescence intensity, while the larger gold nanoparticles (100 nm) offered greater than 60% increase in fluorescence. Our results are in agreement with their observations, with the larger gold nanoparticles offering a greater contrast ratio. When Cy5 was used as the fluorophore, the trend in the contrast ratio as a function of nanoparticle size (Fig. 3b) was similar to that of Cy3. While the trend was similar, the contrast ratios for the 4 and 87 nm gold nanoparticles were 3-fold greater than the 15 nm gold nanoparticles (compared to
